Abstract: A form feature hierarchy recognition method based on knowledge and graph is proposed to solve the problem of available management of relationship among features. Form features are recognized in different layers including Face Relation Graph(FRG), Simple Form Feature Relation Graph(SFFRG) and Compound Form Feature Relation Graph(CFFRG). Data structure of model and feature recognition algorithms are provided. Experiment shows that the method can simplify Attributed Adjacency Graph(AAG) of parts, and rich feature relationship graph with more engineering information. It provides a new way to establish liaison assembly features.

Key words: knowledge base, Attributed Adjacency Graph(AAG), form feature, hierarchy recognition

摘要： 为解决特征识别中对特征间关系缺乏有效处理的问题，提出基于知识与图的形状特征层次识别方法。将形状特征进行分层识别，由低到高依次分为面关系图、简单形状特征关系图、组合形状特征关系图，给出模型的数据结构及特征识别算法。实验证明，该方法可以简化零件的属性邻接图表达方法，丰富图表示法的工程含义，为建立零件间的关联功能级装配特征提供一条新的途径。

关键词: 知识库, 属性邻接图, 形状特征, 层次识别

CLC Number: 

• TP391.41